anth-debug-bundle
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The script includes a command that pipes a JSON response from status.anthropic.com into a Python one-liner. While automated scanners flagged this as potential remote execution, the command uses a hardcoded Python script to parse data from a trusted vendor domain rather than executing arbitrary remote code.
- [DATA_EXFILTRATION]: The skill interacts with the ANTHROPIC_API_KEY environment variable to verify connectivity. The script explicitly avoids logging the secret value, instead using shell expansion to log only the presence and character length of the key for troubleshooting purposes.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to capture external data including API response headers and logs. This creates a surface where external content is ingested into the agent context; however, the skill provides specific guidelines for redacting user content and PII before the data is shared with support.
- [COMMAND_EXECUTION]: The debug script utilizes standard system tools like tar, curl, and grep to aggregate diagnostic information. The tools are scoped appropriately to the skill's intended purpose of generating a support archive.
Audit Metadata